Awareness (accurate, conscious understanding) of one's own cognitive, emotional, and social abilities and limitations is among the highest of all cerebral functions. This aspect of consciousness is often impaired several months to years after a moderate to severe traumatic brain injury (TBI) and often interferes with treatment and adjustment following injury. The neural correlates of deficits in self-awareness are not well understood, and progress in this area may be helpful in predicting recovery and designing optimal treatment strategies. In this application we propose to use BOLD contrast functional MRI (fMRI) and diffusion tensor imaging (DTI) to probe the neural substrates of self-awareness, as well as executive functions and episodic memory in TBI cases and matched trauma controls (orthopedic, non-brain trauma). We will examine the presence and degree of resolution of injury-induced cognitive and psychosocial deficits, and relate these to neural activity using fMRI during cognitive tasks, and to axonal integrity as measured by DTI. We will examine all subjects (60 moderate, 60 severe TBI cases, and 60 matched trauma controls) at 4 months post-injury, and again at 18 months in order to test the following hypotheses: 1. The anterior medial prefrontal cortex subserves self-reflective thought. Activation in this region will be related to the patient's level of self-awareness, and to the degree of white matter integrity underlying that region. 2. Improvement of executive and memory abilities will be related to longitudinal changes in activation in the lateral prefrontal cortex and mesial temporal cortex respectively. Improvement in cognition, together with a longitudinal change in activation, and a difference from control group activity, will be considered evidence for cerebral reorganization. 3. Functional imaging findings will improve prediction of neurobehavioral outcome. Approach. We will characterize patients with neuropsychological and neurological evaluation, neurobehavioral outcomes, and functional MR imaging. Depth and duration of coma will be used as primary measures of injury severity. MRI activation and tensor data will be analyzed using the general linear model with a voxel-based approach. This technique will allow direct and simultaneous assessment of the relationship between MR data and other measurements such as injury severity, outcome and scores on cognitive tests. We expect that successful completion of this project will provide greater understanding of patients with debilitating brain injury, and that these probes of neurobehavioral function may be useful in the future to guide and monitor treatment interventions.